Clinical Approach to Fever of Unknown Origin in Children Ya-Li Chien A,B, Fang-Liang Huang A, Chung-Ming Huang A, Po-Yen Chen A,*

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

+ MODEL provided by Elsevier - Publisher Connector Journal of Microbiology, Immunology and Infection (2015) xx,1e6

Available online at www.sciencedirect.com ScienceDirect

journal homepage: www.e-jmii.com

ORIGINAL ARTICLE Clinical approach to fever of unknown origin in children Ya-Li Chien a,b, Fang-Liang Huang a, Chung-Ming Huang a, Po-Yen Chen a,*

a Section of Pediatric Infectious Disease, Department of Pediatrics, Taichung Veterans General Hospital, Taichung, Taiwan b Department of Pediatrics, Lin Shin Medical Corporation Lin Shin Hospital, Taichung, Taiwan

Received 29 April 2015; received in revised form 7 August 2015; accepted 20 August 2015

Available online ---

KEYWORDS Abstract Background/purpose: Fever of unknown origin (FUO) can be caused by many clin- children; ical conditions and remains a diagnostic challenge in clinical practice. The etiology of FUO var- fever of unknown ies markedly among different age groups, geographic areas, and seasons. A four-stage origin investigative protocol for FUO is widely applied in clinical practice. The aim of this study was to evaluate the usefulness of this four-stage protocol for identifying the etiology of FUO in children. Methods: We enrolled children younger than 18 years of age who were admitted to the Tai- chung Veterans General Hospital during the period from January 2006 to December 2014 with FUO persisting for more than 3 weeks. The four-stage FUO investigative guideline was used to evaluate the etiology of fever in all patients enrolled in the study. Results: The etiology of FUO was identiﬁed in 79 (84.9%) of the 93 patients enrolled in the study. The most common cause of FUO was infectious disease (37.6%), followed by malignancy (17.2%), miscellaneous disease (16.1%), and collagen vascular disease (14.0%). With respect to the four-stage survey of FUO, 36 of the 79 patients (45.6%) were identiﬁed in Stage 3, 28 patients (35.4%) in Stage 2, 13 patients (16.5%) in Stage 4, and only two patients (2.5%) in Stage 1. Conclusion: A well-designed systemic review of the epidemiological information, medical history, physical examination, laboratory analysis, and adequate invasive procedures provide adequate data to identify the most common causes of FUO in children. Copyright ª 2016, Taiwan Society of Microbiology. Published by Elsevier Taiwan LLC. All rights reserved.

* Corresponding author. Division of Infection, Department of Pediatrics, Taichung Veterans General Hospital, Number 1650, Section 4, Taiwan Boulevard, Taichung 407, Taiwan. E-mail address: [email protected] (P.-Y. Chen).

Please cite this article in press as: Chien Y-L, et al., Clinical approach to fever of unknown origin in children, Journal of Microbiology, Immunology and Infection (2015), http://dx.doi.org/10.1016/j.jmii.2015.08.007 + MODEL 2 Y.-L. Chien et al.

Introduction environmental factors, and vaccinations, and conducting a more detailed physical examination and assessment of Fever is a common symptom of many clinical conditions, fever pattern, specific screening tests, and noninvasive and infection is the most common cause, especially in diagnostic tests, including cultures for viruses, serological e children.1 “Fever of unknown origin (FUO)” in adults was tests for bacteria and viruses (Weil Felix test, Widal test, first described in 1961 and was defined as well-documented and serological tests for leptospiral and herpes group vi- fever of at least 3 weeks’ duration without an apparent ruses), antinuclear antibody testing, complements 3 and 4 source after 1 week of investigation.2 Although there is no level, testing for cytoplasmic antineutrophil cytoplasmic standard definition of pediatric FUO, fever lasting any- antibodies, measurement of immunoglobulin level, and ul- where from 10 days to 3 weeks is generally accepted as the trasonographic examination of tissue or organs (abdominal, working definition of FUO in children.3e6 pelvic, chest, and lymph node). Stage 3 of the protocol FUO is a challenging problem in clinical practice, and involves the undertaking of special diagnostic tests (for Q patients who present with FUO are often subjected to un- fever, Lyme disease and Parvovirus, etc.), which can be necessary over-the-counter laboratory tests and antimi- conducted at the reference laboratories of the Centers for crobial therapies. There is no gold “standard” process to Disease Control, R.O.C. (Taiwan), radiological examinations investigate the etiology of FUO. There are only a limited (computed tomography scan, magnetic resonance imaging, number of reports on systemic surveys of the etiology of and magnetic resonance angiography), nuclear medicine FUO among children in Taiwan. In this study, we investi- (Gallium scan and bone scan), and tissue or bone marrow gated the four-stage protocol for FUO,7,8 and evaluated biopsy. Further specific diagnostic examination includes which stage provides the most valuable information for echocardiography, polymerase chain reaction for HIV and identifying the cause of FUO in children. cytomegalovirus (CMV), and molecular DNA analysis for Fabry’s disease. Stage 4 investigation for FUO includes therapeutic trials including empiric antimicrobial therapy Methods and administration of corticosteroids, intravenous immu- noglobulins, and prostaglandin inhibitors. 2 According to the classical Petersdorf and Beeson criteria, The identified causes of FUO in children were divided the working definition of FUO used in this study is well- into the following five groups: infectious disease, collagen documented fever (>38.3 C) of at least 3 weeks’ duration vascular disease, malignancy, miscellaneous conditions, without an apparent source after 3 weeks of evaluation as and undiagnosed. Patients in whom the etiology of fever an outpatient or after a week of evaluation in hospital. was not identified and who recovered spontaneously were Children younger than 18 years of age who presented to assigned to the undiagnosed group.9 the Department of Pediatrics at the Taichung Veterans All statistical analyses were performed using the Statis- General Hospital with prolonged or unrecognized source of tical Package for Social Science (SPSS Inc., Chicago, IL, fever during the period from January 2006 to December USA). 2014 and who fulfilled the FUO definition for children were enrolled in this study. Exclusion criteria included any previous immunocompromised disease state [e.g., human im- Results munodeficiency virus (HIV) infection, leukemia, lymphoma] or chronic medical condition. A total of 93 children were enrolled in this study, including Children with FUO who fulfilled the inclusion criteria 58 boys (62.4%) and 35 girls (37.6%). The mean age was were evaluated systemically based on the four-stage FUO 7.3 Æ 5.5 years (range 2 mo to 18 y). Nine children (9.7%) protocol.7,8 In brief, Stage 1 comprises the taking of a were younger than 1 year of age, 31 (33.3%) were in the age detailed medical history, examination of medical records, a range from 1 year to 5 years, 31 (33.3%) were in the age physical examination, and performance of screening tests, range from 6 years to 12 years, and 22 (23.7%) were in the including measuring complete blood count/differential age range from 12 years to 18 years. Infectious diseases count, C-reactive protein level, examination of peripheral were the most common cause of FUO (n Z 35, 37.6%), fol- blood smear, urinalysis, routine biochemical analysis, cul- lowed by malignancies (n Z 16, 17.2%), miscellaneous dis- ture of urine, stool and blood sample, and radiographic eases (n Z 15, 16.1%), and collagen vascular diseases examination of chest and sinus cavities. Stage 2 involves (n Z 13, 14.0%). Systemic investigation did not yield a collecting a more detailed history of travel, contact, diagnosis in 14 (15.1%) patients (Table 1).

Table 1 Age distribution of different categories of fever of unknown origin in children. Cause <1y 1e5y 6e12 y >12 y Total Infectious disease 7 (77.8) 8 (25.8) 12 (38.7) 8 (36.4) 35 (37.6) Malignancy 0 (0) 7 (22.6) 6 (19.4) 3 (13.6) 16 (17.2) Collagen vascular disease 0 (0) 4 (12.9) 5 (16.1) 4 (18.2) 13 (14.0) Miscellaneous 1 (11.1) 4 (12.9) 6 (19.4) 4 (18.2) 15 (16.1) Undiagnosed 1 (11.1) 8 (25.8) 2 (6.4) 3 (13.6) 14 (15.1) Total 9 (9.7) 31 (33.3) 31 (33.3) 22 (23.7) 93 (100) Data presented as n (%).

The mean age of patients with FUO caused by infectious prostaglandin inhibitor or corticosteroids, and in seven of disease, malignancy, collagen vascular diseases, and the patients FUO was laterally attributed to juvenile miscellaneous diseases was 7.0 Æ 5.4 years, 8.9 Æ 5.2 years, rheumatoid arthritis (JRA; Table 3). 6.5 Æ 4.3 years, and 8.5 Æ 6.4 years, respectively. The mean The identified causes of FUO in this study are listed in age of children for whom the etiology could not be deter- Table 4. Leptospirosis (4 patients) and rickettsial disease (4 mined was 5.6 Æ 6.2 years. The most common cause of FUO patients) were the most common infectious diseases in in children aged younger than 1 year was infectious disease children with FUO. Nonspecific presentations and inade- (77.8%). All patients with FUO due to collagen vascular quate history taking, physical examination, and empirical disease or malignant disease in this study were older than 1 antimicrobial therapies were the most common reasons for year of age. The age of children for whom the etiology delay in diagnosis. could not be determined mostly ranged from 1 year to 5 Typhoid fever was the cause of FUO in two patients. The years (Table 1). mean duration of fever in those patients was 3 weeks. They Following the protocol of systemic investigation for FUO, were aged 1 year and 2 years and neither patient presented 36 of the 79 patients (45.6%) with FUO were identified in with relative bradycardia or rash. The laboratory studies Stage 3, 28 patients (35.4%) in Stage 2, 13 patients (16.5%) were all nonspecific. Blood culture was negative for the in Stage 4, and two patients (2.5%) in Stage 1 (Table 2). For organisms; however, the results of the Widal test were those patients with FUO caused by infectious diseases, the confirmative of typhoid fever. Results of echocardiographic etiologies for 19 patients (54.3%) were found in Stage 2 by examinations and blood cultures revealed infective endo- reviewing additional medical history, careful physical ex- carditis in two patients (Staphylococcus aureus in one and amination, and common epidemiological screening testing Streptococcus parasanguinis in the other). [WeileFelix test for Rickettsia disease, Widal test for Sal- Of the malignant diseases causing FUO, acute leukemia monella, Leptospiral test, and EpsteineBarr virus (EBV) was the most common, followed by neuroblastoma and antibodies response, etc.]. All children with malignancy lymphoma. Of the collagen vascular diseases causing FUO, needed an invasive procedure (bone marrow or tissue bi- JRA was the most common, followed by systemic lupus opsy) to make the diagnosis in Stage 3. Tissue biopsy in erythematosus, Sjo¨gren’s syndrome, and sarcoidosis. The Stage 3 was also useful for identifying hemophagocytosis miscellaneous causes of FUO among children included Ka- and KikuchieFujimoto lymphadenitis in the miscellaneous wasaki disease (KD) and KikuchieFujimoto lymphadenitis. group. Eight patients (61.5%) with collagen vascular disease A definitive diagnosis was not achieved using the four- were identified in Stage 4 using the therapeutic trials of stage protocol in 14 of the 93 patients. Eleven of them were

Table 2 Etiologies of fever of unknown origin identified by the programmed four-stage investigation procedure among 79 children in Taichung Veterans General Hospital. Stage Investigation procedure Measurement n % 1 History, physical examination, CBC/DC, CRP, peripheral blood smear, complete urine 2 2.5 screening test analysis, routine biochemistry, chest/sinus radiography, urine/stool/blood culture 2 More detailed history, repeat Virus culture, ANA, C3/C4, C-ANCA, serological test 28 35.4 physical examination, specific for bacteria and virus, immunoglobulin level, screening test abdominal/chest/lymph node ultrasonography 3 Specific diagnostic test, CT, MRI, gallium scan, bone marrow aspiration and 36 45.6 invasive test biopsy, biopsy of lymph nodes/mass, PCR for CMV and HIV, echocardiography, lumbar puncture, molecular DNA analysis for Fabry’s disease 4 Therapeutic trial Antimicrobial therapy, NSAID/steroid trial 13 16.5 ANA Z antinuclear antibody; C3/C4 Z complements 3 and 4; C-ANCA Z cytoplasmic antineutrophil cytoplasmic antibodies; CBC/ DC Z complete blood count/differential count; CMV Z cytomegalovirus; CRP Z C-reactive protein; CT Z computerized tomography; HIV Z human immunodeficiency virus; MRI Z magnetic resonance imaging; NSAID Z nonsteroidal anti-inflammatory drug; PCR Z polymerase chain reaction.

Table 3 Value of the four-stage investigation procedure to establish the etiology of fever of unknown origin in children. Etiology (n) Stage of diagnosis, n (%) 12 3 4 Infectious disease (35) 2 (5.7) 19 (54.3) 9 (25.7) 5 (14.3) Malignancy (16) 0 (0.0) 0 (0.0) 16 (100.0) 0 (0.0) Collagen vascular disease (13) 0 (0.0) 5 (38.5) 0 (0.0) 8 (61.5) Miscellaneous (15) 0 (0.0) 4 (26.7) 11 (73.3) 0 (0.0)

Table 4 Identified causes of fever of unknown origin in children. Etiology (n) Diagnosis (n) n Infectious disease (35) Viral (9) HIV 1 Coxsackievirus B4 1 CMV 2 Parvovirus B19 1 HSV type 1 1 Parainfluenza type 3 1 Influenza type B 1 EBV 1 Bacterial (26) Infective endocarditis 2 TB meningitis 1 TB lymphadenopathy 1 Leptospirosis 4 Rickettsiae disease 4 Typhoid fever 2 Pulmonary actinomycosis 1 Toxoplasmosis 1 Pneumonia 1 Pharyngotonsillitis 2 Acute otitis media 1 Acute sinusitis 1 Osteomyelitis 1 Perivertebral abscess and diskitis 1 Atypical infection 3 Malignancy (16) Acute lymphoblastic leukemia 9 Acute myelogenous leukemia 1 Anaplastic large cell lymphoma 2 Peripheral T-cell lymphoma 1 Neuroblastoma 3 Collagen vascular disease (13) Juvenile rheumatoid arthritis 8 Sjo¨gren’s syndrome 1 Sarcoidosis 1 SLE 1 Undifferentiated connective tissue disease 2 Miscellaneous (15) KikuchieFujimoto lymphadenitis 4 Kawasaki disease 5 Hemophagocytosis 2 Fabry’s disease 1 Hypohidrotic ectodermal dysplasia 1 Lymphadenopathy 1 Hypogammaglobulinemia 1 Undiagnosed 14 Total 93 CMV Z cytomegalovirus; EBV Z EpsteineBarr virus; HIV Z human immunodeficiency virus; HSV Z herpes simplex virus; SLE Z systemic lupus erythematosus; TB Z Mycobacterium tuberculosis.

afebrile without any sequelae at follow up. The other three frustration for the attending physician, and leads to un- patients remained febrile but were lost to follow up. necessary and additional over-the-counter laboratory tests Four patients died during the study period. One of them and medications (including antimicrobial agents), subse- died due to the complication of infective endocarditis, two quently increasing the burden on both the National Health died due to neuroblastoma, and one died due to acute Insurance system and the patient’s family. A well-designed leukemia. systemic investigation protocol conducted according to local epidemiological information may save time and reduce unnecessary medical costs and disease sequelae. Discussion There are few available data on children with FUO in Taiwan. Based on the four-stage procedure to investigate Fever in pediatric patients causes considerable anxiety to the etiology of FUO among children, most cases were their parents, while FUO is a source of confusion and identiﬁed in Stage 2.

In this study, we found that the most common cause of presented in 44e86% of patients with scrub typhus.22,23 In FUO was infectious disease (37.6%) followed by malignancy, our study, approximately 25% of patients with scrub typhus miscellaneous diseases, and collagen tissue disease. Similar presented with eschar. Rose spots are commonly seen in findings have been reported previously.10e13 A recent study patients with typhoid fever24 and conjunctival suffusion is of children with FUO conducted by Cho et al14 in northern commonly noted in patients with leptospirosis.20 Limping Taiwan, in which FUO was defined as fever for more than 2 gait and sternal tenderness may indicate the presence of weeks, found that infectious diseases accounted for 27% of malignant invasion of the bone marrow.25 FUO in children, followed by malignancies (16.6%), and Serological tests for atypical bacteria (WeileFelix test collagen vascular disease (12.7%). Cho et al14 also found and leptospiral antibody) have been shown to have low that upper respiratory infections (URIs) were the most sensitivity and specificity. Leptospirosis is an important common cause of infectious disease, followed by CMV and endemic zoonotic disease in Taiwan. Culture and micro- EBV infections; leptospirosis or scrub typhus rarely caused scopic agglutination tests are the gold standard methods for FUO in their study. There were more urban cases in Taipei establishing a diagnosis of leptospirosis; however, they are (Northern Taiwan), whereas there were more rural cases in only available from reference laboratories of the Centers Central Taiwan, and the criterion for duration of fever in for Disease Control, R.O.C (Taiwan). Whole-cell-based northern and central Taiwan was 2 weeks and 3 weeks, rapid serological tests have been shown to have a low respectively. Viral URI usually caused fever for 3e5 days, sensitivity for early phase leptospirosis and a low specificity and was rarely longer than 10 days, except in cases with in highly endemic areas.26 For patients with high clinical adenovirus/influenza infection or complication with otitis suspicion of leptospirosis, empirical therapeutic trial should media or bronchopneumonia/pneumonia. CMV and EBV in- be initiated (Stage 4). fections in children usually presented with mononucleosis- Investigation studies for FUO in Stage 3 include invasive like syndrome15; they were easily reactivated in patients procedure. Clinical suspicion of acute leukemia/lymphoma with immunocompromised status,16 and their possible roles and malignancy always requires bone marrow studies. in prolonged fever should be carefully interpreted. Chin There were four cases of KikuchieFujimoto disease in this et al17 reported that infection was the most common eti- study, and all were definitely diagnosed by lymph node ology of FUO in adults in Taiwan, which was similar to the biopsy.27 rate in children. Tuberculosis was the most common in- Infective endocarditis is commonly found in pediatric fectious etiology in their study. That finding is compatible patients.28 Echocardiography should be arranged based on with the higher annual incidence rate of tuberculosis in clinical suspicion, especially for those with positive blood adults (3.04/100,000 population and 82.60/100,000 popu- cultures.29 Echocardiography is also very helpful when KD is lation for children and adults, respectively) reported by suspected, even in young children with FUO who did not Tsai et al.18 meet the typical criteria of KD.30,31 In five KD patients in Malignancies were previously found to be the third most this study, four were diagnosed with incomplete-type KD, common cause of FUO in children, following infectious and based on echocardiographic finding and laboratory collagen vascular disease. However, malignancies were characteristics.32 slightly more common than collagen vascular diseases in Gallium scanning is capable of identifying inflammatory this study. Most cases of FUO due to collagen vascular dis- lesions and tumors throughout the entire body. However, eases had been intensively surveyed in a local hospital, and this modality was not commonly used in our study. A pre- were referred to medical centers for further investigation. vious study also found that gallium scanning was rarely The percentage of FUO in children who remained unrec- useful in most children with FUO.33 Magnetic resonance ognized was similar to previously reported rates, which imaging was the imaging modality of choice for establishing ranged from 12% to 24%.10e13 the diagnosis of osteomyelitis and arthritis or to delineate The four-stage protocol used in this study to investigate the location and extent of bone and soft-tissue the etiology of FUO provides for a reasonably accurate involvement.34,35 establishment of a diagnosis. Stage 2 of the protocol was Therapeutic trial with anti-inflammatory medications or sufficient to establish a diagnosis of infectious disease as antibiotics should be generally avoided as empirically the cause of FUO in most patients in this study. Clinicians diagnostic measures in children with FUO, except for chil- who encounter children with FUO need to collect a detailed dren with suspected JRA for whom nonsteroidal agents are record of disease and medical history, including vaccination administered, but this is a diagnosis of exclusion.36 How- history, growth and developmental milestones, and body ever, it is always delayed since the onset of fever and the weight changes. Cough contact history with elders with TB progression to arthritis. All the cases of JRA diagnosed in is an indicator for tuberculosis, and poor weight gain or our study had a good response to treatment with nonste- failure to thrive is suggestive of an autoimmune or immu- roidal anti-inflammatory drugs. nodeficiency disorder. Some common endemic infectious Empirical treatment with broad-spectrum antibiotics diseases (leptospirosis and rickettsial diseases) are easily will mask or delay the diagnosis of some infectious dis- overlooked in clinical practice.19e21 A detailed evaluation eases; however, broad-spectrum antibiotics are the most of travel, occupation, contact, and cluster may help to commonly applied medication in children to treat the establish whether the patient lives in a rural area or has infection. Diseases that cannot be routinely tested include recently visited a mountainous area. Lyme disease, leptospirosis, and nonscrub typhus rickettsial A more detailed physical examination is needed for all diseases. Empirical antimicrobial therapy after blood/tis- patients with FUO. Skin rashes or lesions are important sue sampling for these diseases may provide a diagnostic clues for diagnosing some infectious diseases. Eschar is clue.

In conclusion, children with FUO should be evaluated 17. Chin C, Chen YS, Lee SS, Wann SR, Lin HH, Lin WR, et al. Fever using the four-stage protocol. Epidemiological data, con- of unknown origin in Taiwan. Infection 2006;34:75e80. tact history, previous medical history, fever pattern, 18. Tsai KS, Chang HL, Chien ST, Chen KL, Chen KH, Mai MH, et al. detailed physical examination, and screening tests provide Childhood tuberculosis: epidemiology, diagnosis, treatment, e adequate information to establish a diagnosis. Stage 3 and vaccination. Pediatr Neonatol 2013;54:295 302. 19. Abgueguen P, Delbos V, Blanvillain J, Chennebault JM, Cottin J, procedures should be reserved for special consideration. Fanello S, et al. Clinical aspects and prognostic factors of leptospirosis in adults. Retrospective study in France. J Infect Conflicts of interest 2008;57:171e8. 20. Vanasco NB, Schmeling MF, Lottersberger J, Costa F, Ko AI, None declared. Tarabla HD. Clinical characteristics and risk factors of human leptospirosis in Argentina (1999e2005). Acta Trop 2008;107: 255e8. References 21. Yang HY, Hsu PY, Pan MJ, Wu MS, Lee CH, Yu CC, et al. Clinical distinction and evaluation of leptospirosis in Taiwanda case- e 1. Finkelstein JA, Christiansen CL, Platt R. Fever in pediatric control study. J Nephrol 2005;18:45 53. primary care: occurrence, management, and outcomes. Pedi- 22. Berman SJ, Kundin WD. Scrub typhus in South Vietnam. A study e atrics 2000;105:260e6. of 87 cases. Ann Intern Med 1973;79:26 30. 2. Petersdorf RG, Beeson PB. Fever of unexplained origin: report 23. Sheehy TW, Hazlett D, Turk RE. Scrub typhus. A comparison of on 100 cases. Medicine (Baltimore) 1961;40:1e30. chloramphenicol and tetracycline in its treatment. Arch Intern e 3. Feigin RD. Feigin & Cherry’s textbook of pediatric infectious Med 1973;132:77 80. diseases. 6th ed. Philadelphia, PA: Saunder/Elsevier; 2009. 24. Parry CM, Hien TT, Dougan G, White NJ, Farrar JJ. Typhoid e 4. Seashore CJ, Lohr JA. Fever of unknown origin in children. fever. N Engl J Med 2002;347:1770 82. Pediatr Ann 2011;40:26e30. 25. Jonsson OG, Sartain P, Ducore JM, Buchanan GR. Bone pain as 5. Chow A, Robinson JL. Fever of unknown origin in children: a an initial symptom of childhood acute lymphoblastic leukemia: systematic review. World J Pediatr 2011;7:5e10. association with nearly normal hematologic indexes. J Pediatr e 6. Mahmoudi S, Mehrazmay A, Salesi M, Mamishi S. Fever of un- 1990;117:233 7. known origin: a retrospective study of 95 children in an Iranian 26. Toyokawa T, Ohnishi M, Koizumi N. Diagnosis of acute lepto- e referral hospital. Br J Biomed Sci 2014;71:40e2. spirosis. Expert Rev Anti Infect Ther 2011;9:111 21. 7. Hassan RH, Fouda AE, Kandil SM. Fever of unknown origin in 27. Kung IT, Ng WF, Yuen RW, Chan JK. Kikuchi’s histiocytic children: a 6-year experience in a tertiary pediatric Egyptian necrotizing lymphadenitis. Diagnosis by fine needle aspiration. e hospital. Int J Health Sci (Qassim) 2014;8:13e9. Acta Cytol 1990;34:323 8. 8. Mims C, Playfair J, Roitt I, Wakelin D, Williams R. Pyrexia of 28. Murdoch DR, Corey GR, Hoen B, Miro JM, Fowler Jr VG, Bayer AS, et al. Clinical presentation, etiology, and outcome of unknown origin. In: Medical microbiology. 2nd ed. New York: st Mosby; 1998. p. 381e5. infective endocarditis in the 21 century: the International 9. Kejariwal D, Sarkar N, Chakraborti SK, Agarwal V, Roy S. Py- Collaboration on Endocarditis-Prospective Cohort Study. Arch e rexia of unknown origin: a prospective study of 100 cases. J Intern Med 2009;169:463 73. Postgrad Med 2001;47:104e7. 29. Hoen B, Duval X. Infective endocarditis. N Engl J Med 2013; 10. Tezer H, Ceyhan M, Kara A, Cengiz AB, Devrim I, Secmeer G. 369:785. Fever of unknown origin in children: the experience of one 30. Burns JC, Glode MP. Kawasaki syndrome. Lancet 2004;364: e center in Turkey. Turk J Pediatr 2012;54:583e9. 533 44. 11. Ciftci E, Ince E, Dogru U. Pyrexia of unknown origin in children: 31. McCrindle BW, Li JS, Minich LL, Colan SD, Atz AM, Takahashi M, a review of 102 patients from Turkey. Ann Trop Paediatr 2003; et al. Coronary artery involvement in children with Kawasaki 23:259e63. disease: risk factors from analysis of serial normalized mea- e 12. Cogulu O, Koturoglu G, Kurugol Z, Ozkinay F, Vardar F, surements. Circulation 2007;116:174 9. Ozkinay C. Evaluation of 80 children with prolonged fever. 32. Cimaz R, Sundel R. Atypical and incomplete Kawasaki disease. e Pediatr Int 2003;45:564e9. Best Pract Res Clin Rheumatol 2009;23:689 97. 13. Pasic S, Minic A, Djuric P, Micic D, Kuzmanovic M, Sarjanovic L, 33. Buonomo C, Treves ST. Gallium scanning in children with fever e et al. Fever of unknown origin in 185 paediatric patients: a of unknown origin. Pediatr Radiol 1993;23:307 10. single-centre experience. Acta Paediatr 2006;95:463e6. 34. Buchmann RF, Jaramillo D. Imaging of articular disorders in e 14. Cho CY, Lai CC, Lee ML, Hsu CL, Chen CJ, Chang LY, et al. children. Radiol Clin North Am 2004;42:151 68. Clinical analysis of fever of unknown origin in children: a 10- 35. Browne LP, Mason EO, Kaplan SL, Cassady CI, Krishnamurthy R, year experience in a northern Taiwan medical center. J Guillerman RP. Optimal imaging strategy for community- Microbiol Immunol Infect 2015 [in press]. acquired Staphylococcus aureus musculoskeletal infections in e 15. Rafailidis PI, Mourtzoukou EG, Varbobitis IC, Falagas ME. Se- children. Pediatr Radiol 2008;38:841 7. vere cytomegalovirus infection in apparently immunocompe- 36. Petty RE, Southwood TR, Manners P, Baum J, Glass DN, tent patients: a systematic review. Virol J 2008;5:47. Goldenberg J, et al. International League of Associations for 16. Ramanan P, Razonable RR. Cytomegalovirus infections in solid Rheumatology classification of juvenile idiopathic arthritis: organ transplantation: a review. Infect Chemother 2013;45: second revision, Edmonton, 2001. J Rheumatol 2004;31: e 260e71. 390 2.